Answers are grounded in this paper and cite their sources.
A new estimation procedure to predict the size Of a design event, given a past flood record, is found to be superior to various alternative fitting procedures. The ‘incomplete means’ method uses no moments higher than the first, a procedure which results in little bias. Although the fitting procedure can be applied to any distribution, it is tested by using a variant of the Wakeby distribution. Using a small Monte Carlo experiment with a single parent distribution, the incomplete means method applied to the logs of the observations results in less regret than other fitting procedures tested. A regret model is developed which introduces a design event based on the density function as an alternative to the more traditional T‐year event.
No takes yet. Share an insight, caveat, or question.
John C. Houghton (1978) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: